The perception of the 3D environment involves several levels of analysis including the perception of the overall depth of a scene, the perception of the layout (relative positions) of objects and surfaces in the scene, and the perception of properties of objects, such as depth and shape. The information at each level may be viewer-centered, indicating absolute or relative distances form the observer, or object-centered, indicating distances relative to objects or to the scene. The proposed research considers the integration of visual information for the perception of 3D sciences according to these two dimensions-level of analysis (scene, layout, and object) and type of information (viewer-centered or object- centered). Specifically, it examines how this integration is affected by interactions of the observer with the scene-determined by the perceptual tasks the observer is asked to perform, the organization of information within the scene for performing these tasks, and whether or not the observer actively explores the scene. The proposed research is divided into four series of experiments. The first series will consider how four types of motion-translation and rotation of an object and of the surrounding scene-determine the perceived depth of a 3D objects in a dynamic scene. The second series will assess the effect of a prior task involving judgements of scene depth, layout, or object depth on subsequent performance of a task at a different level of the hierarchy. The third series will examine how the introduction of uniform regions connecting objects at different depths in a scene affecting judgements about objects and regions in the scene. The fourth series of experiments, to be conducted in a virtual reality laboratory, will examine the effects of active observation (e.g., walking within a scene) on the accuracy of distance judgements, when the optic flow information for active and passive observers is exactly matched. An integrated approach to the study of the perception of 3D scenes is important for the development of perceptual models concerned with the analysis of 3D scenes. Understanding how interactions between the observer and information in the scene affects integration of different types of information (viewer-centered and object- centered) and different levels of analysis (scene, layout and object) should be important for research in areas of visual cognition concerned with 3D objects and scenes, especially research on object recognition and visual attention.